f2fs: schedule in between two continous batch discards
authorChao Yu <yuchao0@huawei.com>
Sun, 21 Aug 2016 15:21:30 +0000 (23:21 +0800)
committerJaegeuk Kim <jaegeuk@kernel.org>
Thu, 8 Sep 2016 00:27:33 +0000 (17:27 -0700)
commit74fa5f3d43bca87257e9da7da95be8735ffa2b96
treea3d972110c11454cbf7c5964bd49184863d9f229
parent97c1794a5dc160164aa7f161310da15c34d62641
f2fs: schedule in between two continous batch discards

In batch discard approach of fstrim will grab/release gc_mutex lock
repeatly, it makes contention of the lock becoming more intensive.

So after one batch discards were issued in checkpoint and the lock
was released, it's better to do schedule() to increase opportunity
of grabbing gc_mutex lock for other competitors.

Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
fs/f2fs/segment.c